YouTube Dashboard — User Manual

The OBS Controller lets you remotely control OBS Studio running on a streaming device at a remote site — start/stop streaming and recording, switch scenes, toggle scene items, manage stream keys, refresh browser overlays, open projectors on specific monitors, and mix audio — all from your browser, without touching the remote machine.

Two Communication Paths

The controller uses two separate communication paths to OBS, depending on the feature:

On the Streaming Device

• OBS Studio must be installed and running.
• WebConnect must be running on the device — this is the bridge that receives commands from the cloud and executes them in OBS. Without it, nothing on this page will work.
• OBS WebSocket Server must be enabled in OBS (Tools → WebSocket Server Settings) for the Audio Mixer to connect. If you set a WebSocket password, it must be saved on the device record in StreamStudio.
• Caddy (or equivalent reverse proxy) must be running with routes for both the WebConnect HTTP path and the /obs-ws WebSocket path, exposed via Tailscale Funnel.

In StreamStudio

• You must be logged in and have been granted access to at least one device by an admin.
• The device must have a Scoreboard ID (used by WebConnect) and a Proxy URL (Tailscale Funnel URL used for the WebSocket connection).

WebConnect Command Flow (HTTP path)

Every command button (except Audio Mixer) follows this three-step flow through the obsCommand backend function:

OBS WebSocket v5 (Audio Mixer path)

The Audio Mixer opens a direct, persistent WebSocket connection to the device's OBS WebSocket server (routed through the Tailscale proxy at <proxy_url>/obs-ws). This provides real-time volume and mute changes with no polling delay, OBS WebSocket v5 authentication (SHA-256 challenge-response if a password is set), and automatic reconnection.

The connection status is shown as a colored dot:

OBS Commands Used

Device Selector

At the top-right of the page, a dropdown lets you select which streaming device to control. Only devices you've been granted access to appear. Your selection is remembered between sessions.

Next to the dropdown, a status indicator shows whether the device is online:

Click the check button to manually ping the device and refresh its status.

Command Panels

Command Output

At the bottom of the page, a collapsible Command Output panel shows the label and result of the last command sent. Click the panel header to expand it and see the full output.

When a command succeeds (and isn't a silent query like listing scenes), a response dialog pops up showing:

The dialog also shows the raw OBS response message (expandable under "OBS Message (debug)"), the response status (true/false), the action taken, and the timestamp when OBS processed the command.

Start Stream

Click the green Start Stream button to begin live streaming in OBS. This sends the StartStream command via WebConnect.

Stop Stream

Click the red Stop Stream button to stop streaming in OBS.

Get Stream Status

Click the grey Get Stream Status button to query OBS for the current streaming state (active/inactive, bitrate, uptime, etc.). The result appears in the response dialog.

Manage Stream Key

Click the Manage Stream Key button to open the Stream Key dialog. Here you can:

• Get Current Key — fetches the stream key currently configured in OBS so you can verify it.
• Set Stream Key — pushes the entered key, type, and server to OBS.

OBS can record locally to disk while also streaming. This is useful for keeping a high-quality archive.

Replay Buffer

The Replay Buffer continuously captures the last few seconds of video, letting you save a replay on demand (commonly used in sports broadcasting for instant replays).

Scenes are the different layouts in your OBS production (e.g., "Pre-match", "Live Game", "Scorebug", "Interview").

Getting the Scene List

Click Get All Scenes & Items to fetch all scenes from OBS. The scene list appears below the button, showing each scene name, the active scene (highlighted green with a "Live" badge), and a Set Active label on non-active scenes.

Switching Scenes

Click any scene name to make it the active program scene. The SetCurrentProgramScene command is sent via WebConnect. While switching, the scene shows a "Switching…" indicator.

Viewing and Toggling Scene Items

Click the chevron (▶) next to a scene name to expand it and load its scene items (individual sources within the scene, like cameras, overlays, images).

Each scene item shows:

• The source name
• A visibility toggle (eye icon) — click to show or hide that item within the scene

When you toggle an item, the HideInputSource or UnHideInputSource command is sent. This is useful for quickly hiding/showing a scoreboard overlay, logo, or camera angle without switching scenes.

Browser sources in OBS are web pages loaded as overlays (e.g., a live scoreboard widget, a lower-third graphic, a chat overlay).

Getting the Overlay List

Click Get All Overlays to fetch all browser sources from OBS. Each browser source appears in a list below the button.

Refreshing a Browser Source

Click the Refresh button next to any browser source to force OBS to reload it without cache (refreshnocache command). This is useful when:

• The overlay's content has changed but OBS is showing a stale version
• The overlay page has errored and needs to be reloaded
• You've updated a remote scoreboard or graphic and want it to appear immediately

Projectors let you output a specific scene or source to a physical monitor or display connected to the streaming device (e.g., a projector at the venue showing the live mix).

Getting the Monitor List

Click Get All Monitors to fetch all connected displays from OBS. Each monitor appears in a list showing its name and index.

Opening a Projector

Click the Projector button next to any monitor to open a full-screen projector output on that display. The OpenVideoMixProjector command is sent with the monitor's name and index.

Saving the Default Monitor

Click a monitor name to select it (highlighted blue), then click Save Monitor to Device to persist it as the default projector target for this device. The saved monitor is shown with a green "Saved:" indicator.

The Audio Mixer provides real-time volume and mute control for all audio sources in OBS. Unlike the other panels, it connects directly to OBS via WebSocket for instant feedback.

Connection Status

A colored dot and label at the top of the panel shows the WebSocket connection status:

Getting Audio Sources

Click Get Audio Sources (enabled only when connected) to fetch all audio inputs from OBS. The panel automatically detects native audio captures (WASAPI, PulseAudio, CoreAudio, ALSA) and potential audio sources (browser sources, VLC, media files, FFmpeg — tested for audio capability).

Each audio source appears as a row with:

• Source name (e.g., "Mic/Aux", "Desktop Audio", "Browser Overlay")
• Mute toggle — click the speaker icon to mute (red) or unmute (blue). Changes apply instantly.
• Volume slider — drag to adjust the volume (0–100%). Changes are debounced and sent in real-time.
• Volume readout — shows the percentage and decibel (dB) value (e.g., 75% · -2.5 dB)

How Volume Changes Work

When you drag the slider:

1. The new volume is sent to OBS immediately via WebSocket (SetInputVolume)
2. The displayed value updates optimistically (no waiting for a response)
3. The dB value is calculated from the multiplier: dB = 20 × log₁₀(volume)
4. If the command fails, the slider reverts to the previous value and an error is shown

Admins can configure the WebConnect communication timeouts on the Settings page:

When to adjust these:

• Increase ping timeout if your device has an unstable internet connection and pings are occasionally delayed.
• Increase poll duration if commands frequently time out (OBS takes more than 5 seconds to respond — common on slow hardware like Raspberry Pi).
• Decrease poll duration if you want faster timeout feedback (at the risk of more false timeouts).

The same page also has toggles for the two background automations (Auto Refresh Stream Statuses, Auto Sync YouTube Streams) — these affect the YouTube dashboard, not the OBS Controller directly.

StreamStudio provides two PTZ (Pan-Tilt-Zoom) control surfaces for ONVIF-compatible cameras at your remote sites. Both let you aim and zoom cameras from your browser, but they serve different use cases:

Both pages communicate with cameras using the ONVIF protocol (an open standard for IP camera control). Commands are sent from the cloud through an ONVIF proxy running at the remote site, which forwards them to the camera over the local network.

On the Camera

• The camera must be ONVIF-compatible (Profile S for streaming, Profile PTZ for movement). Most Hikvision, Dahua, and Axis PTZ cameras support ONVIF out of the box.
• ONVIF must be enabled in the camera's web interface (usually under Network → Advanced → ONVIF or Integration → ONVIF).
• An ONVIF user account must be configured on the camera with PTZ control permissions. The credentials are stored in the ONVIF proxy's environment, not in StreamStudio.
• The camera must have a fixed IP address on the local network at the remote site.

On the Remote Site (Raspberry Pi / Windows PC)

• An ONVIF proxy must be running (typically on port 3000) to translate cloud commands into ONVIF SOAP requests to the camera.
• For the Live Camera Feed (Linux only): FFmpeg + mjpg-streamer must be running to convert the camera's RTSP stream into an MJPEG feed served over HTTP.
• Caddy (or equivalent reverse proxy) must expose both the ONVIF proxy and the mjpg-streamer via Tailscale Funnel.

In StreamStudio

• You must be logged in and have been granted access to at least one device by an admin.
• The device must have a Camera IP (the camera's local IP at the site) and a Proxy URL (the Tailscale Funnel URL of the ONVIF proxy).
• For the Live Camera Feed, the device must also have a Stream URL (the Tailscale Funnel URL of the mjpg-streamer) and the OS Type must be set to linux.

All PTZ movement commands go through the ptzCommand backend function, which forwards them to the ONVIF proxy at the remote site. The proxy then sends an ONVIF SOAP request to the camera.

Press-and-Hold Behaviour

Pan, tilt, and zoom buttons use press-and-hold semantics. When you press a direction button, a continuous-move command is sent — the camera starts moving. When you release the button (pointer up or pointer leave), a stop command is sent immediately. This mimics a physical joystick.

Speed

The speed slider (0.1–1.0) controls how fast the camera moves. Lower values are for fine adjustments; higher values are for quick repositioning. The speed is included in every move command sent to the camera.

The PTZ Camera Control page (/ptz) is a focused control surface for camera movement — no video feed, just the controls. It's ideal when you already have a separate video monitor and just need to aim the camera.

Device Selector

At the top-right, a dropdown lets you pick which device's camera to control. Only devices you've been granted access to appear. A status indicator shows whether the camera is online (green) or offline (red). Click the check button to ping the camera via the ONVIF proxy.

If the selected device has no camera IP configured, a warning message appears: "No camera configured — Ask an admin to set a camera IP for this device."

Layout

The page is a single centered card containing three sections:

A speed slider sits at the bottom of the card, and the camera's IP address is shown below it.

The pan-and-tilt pad is a 3×3 grid of buttons. The eight outer buttons move the camera in eight directions; the center button stops all movement.

How to use it

• Press and hold a direction button — the camera starts moving in that direction at the current speed.
• Release the button — a stop command is sent and the camera halts.
• If you drag your finger/cursor off the button, the stop command fires automatically (pointer-leave).
• The center red button sends an immediate stop — useful if the camera is drifting after a release.

Zoom In / Zoom Out

Two buttons below the pan-and-tilt pad control the camera's optical zoom. Like the directional buttons, they use press-and-hold semantics:

• Press and hold Zoom In — the lens zooms in continuously at the current speed.
• Press and hold Zoom Out — the lens zooms out continuously.
• Release — a stop command halts the zoom immediately.

Park (Home)

The Park button sends a home command that returns the camera to its saved home preset position. This is useful for:

• Resetting the camera to a known position before or after a broadcast
• Quickly re-centering on a default framing
• Recovering if the camera has been moved to an unknown position

A speed slider at the bottom of the control card sets how fast the camera moves for all pan, tilt, and zoom commands.

When to adjust:

• 0.1 – 0.3 — fine adjustments, framing a shot precisely while watching the live feed.
• 0.4 – 0.7 — general repositioning, moving between subjects.
• 0.8 – 1.0 — quick pans across a wide area, repositioning for a completely different angle.

The Live Camera Feed page (/local-ptz) combines a live video preview with PTZ controls, image adjustment, and diagnostics. It's the full-featured camera control surface.

Two-Column Layout

The page is split into two columns:

Below the two columns, the Image Adjustment panel and the Diagnostics panel appear full-width.

Windows Devices

The live video feed is an MJPEG (Motion JPEG) stream served by mjpg-streamer on the remote device, which captures frames from the camera's RTSP stream via FFmpeg and serves them over HTTP through Caddy and the Tailscale Funnel.

Connection States

A status indicator in the video panel header shows the current feed state:

How the stream works

Unlike a plain <img> tag, the feed uses the Fetch API with a ReadableStream to decode each JPEG frame individually. This provides true per-frame timing so a silent mid-stream stall is detected and the connection self-heals:

• If no bytes arrive for 10 seconds, the watchdog aborts the connection and reconnects automatically.
• If the Fetch API fails (e.g., CORS blocked), it falls back to a plain <img> tag that displays the stream without frame-level monitoring.
• If the fallback <img> also fails, the feed shows an error state.

Refresh Button

Click Refresh in the video panel footer to force a reconnection. This aborts the current stream connection and starts a fresh one — useful if the feed has frozen or the camera was restarted.

The Image Adjustment panel (Linux devices only) lets you remotely adjust the camera's image settings — brightness, contrast, saturation, and sharpness — via the ONVIF imaging service. Changes apply directly to the camera's image sensor.

Settings

How to use it

1. The panel auto-loads the camera's current settings when you select a device.
2. Drag any slider or type a value in the number field to adjust.
3. Click Apply to push the new values to the camera. A green check appears briefly on success.
4. Click Reset to restore the camera's factory defaults (all values return to 50).

The Diagnostics panel (on the Live Camera Feed page) runs a 3-layer connectivity test to help pinpoint where a connection problem lies — browser, proxy, or camera.

Running Diagnostics

Click Run Diagnostics to start the test. The panel auto-expands and shows a timestamped log of each step with a colored severity level (info, success, warn, error).

The 3 Layers Tested

Common diagnostic outcomes

• Layer 1 returns 502: mjpg-streamer is down — restart it on the device.
• Layer 1 CORS error: Caddy is reachable but missing the Access-Control-Allow-Origin header — fix the Caddyfile.
• Layer 1 network error: Caddy or Tailscale Funnel is down — check both services.
• Layer 2 error: ONVIF proxy is down or the camera credentials in the proxy's .env are wrong.